An increase in the number of predictive factors augments the likelihood of malignancy in branch duct intraductal papillary mucinous neoplasm of the pancreas.

BACKGROUND

International consensus guidelines for the management of intraductal papillary mucinous neoplasms (IPMNs) of the pancreas provide several factors that can be used to predict which IPMNs will become malignant.The sensitivity of each factor's predictive accuracy, however, is relatively low, making it difficult to determine the appropriate treatment in individual cases. The aim of this study was to investigate whether increasing the number of predictive factors might augment the sensitivity of the established guidelines to detect malignant IPMNs.

METHODS

The medical records of 138 patients with IPMNs resected at our institution were reviewed. Possible malignant predictors were analyzed by univariate and multivariate analysis, and the effects of the number of factors and the predictive score of the pathologic results were examined. The cutoff points for the number of predictors to discriminate between malignant and nonmalignant IPMNs were established by constructing receiver operating characteristic curves.

RESULTS

A predictive analysis could not be carried out for the main duct IPMNs because of the high prevalence of malignancy and the small number of significant predictors associated with them. For malignant branch duct IPMNs, however, we identified 4 predictive factors that helped determine the correct diagnosis as follows: (1) the presence of a cyst ≥30 mm in diameter; (2) the presence of mural nodules; (3) a history of acute pancreatitis; and (4) atypical results of pancreatic juice cytology. An increase in the number of these factors significantly affected the sensitivity to predict malignancy. The area under the curve for the number of predictors for malignant branch duct IPMNs was 0.856, and the sensitivity and specificity were 96% and 71%, respectively, when the cutoff point was set at 2. The predictive scoring system also showed the same values of sensitivity and specificity for the number of factors.

CONCLUSION

Patients with branch duct IPMNs who have 2 or more of the 4 predictive factors described above should undergo standard pancreatectomy with lymph node dissection, whereas patients who present with 0 or 1 predictive factor can be treated by minimal pancreatectomy without nodal dissection or by careful observation without resection. All patients with main duct IPMNs, therefore, should be treated with resection as suspected malignancies.